Truncated SVD of the whitened matrix

Computes singular values (optionally truncated) of the whitened matrix. Uses RSpectra when available for efficiency in the truncated case.

Usage

whitened_svd(X, A, M, k = NULL)

Arguments

X, A, M

As in whitened_matrix().

k

Optional integer number of singular values to compute.

Value

A list with d (singular values) and u,v set to NULL.

See also

whitened_matrix() for the underlying whitening operation

Other subspace analysis: procrustes_distance(), subspace_principal_angles(), whitened_matrix()

Examples

# \donttest{
set.seed(1)
X <- matrix(rnorm(300), nrow = 30, ncol = 10)
A <- Matrix::Diagonal(ncol(X))
M <- Matrix::Diagonal(nrow(X))
ws <- whitened_svd(X, A, M, k = 5)
# }